Pancreatic duct glands (PDGs) are a progenitor compartment responsible for pancreatic ductal epithelial repair

Abstract

Pancreatic duct glands (PDGs) have molecular features known to mark stem cell niches, but their function remains to be determined. To investigate the role of PDGs as a progenitor niche, PDGs were analyzed in both humans and mice. Cells were characterized by immunohistochemistry and microarray analysis. In vivo proliferative activity and migration of PDG cells were evaluated using a BrdU tag-and-chase strategy in a mouse model of pancreatitis. In vitro migration assays were used to determine the role of trefoil factor (TFF) -1 and 2 in cell migration. Proliferative activity in the pancreatic epithelium in response to inflammatory injury is identified principally within the PDG compartment. These proliferating cells then migrate out of the PDG compartment to populate the pancreatic duct. Most of the pancreatic epithelial migration occurs within 5days and relies, in part, on TFF-1 and -2. After migration, PDG cells lose their PDG-specific markers and gain a more mature pancreatic ductal phenotype. Expression analysis of the PDG epithelium reveals enrichment of embryonic and stem cell pathways. These results suggest that PDGs are an epithelial progenitor compartment that gives rise to mature differentiated progeny that migrate to the pancreatic duct. Thus PDGs are a progenitor niche important for pancreatic epithelial regeneration.

Figure 1. Tag and chase reveals a dynamic shift of PDG cells

(A) molecular characteristics of PDG. MUC6 is expressed only in PDG (arrow) but not in ductal epithelium (arrowhead), while KRT7 expressed in ductal epithelium but not in PDG. (B) PDG (arrow) is the principal site of proliferation. In response to injury, there is a marked up-regulation of BrdU incorporation and increased expression of the proliferative marker Ki-67. (C) Scheme of BrdU tag-and-chase experiment. BrdU was injected to mark proliferating cells on day 0, and mouse pancreata were harvested daily over the chase period of 7 days. (D) Qualitative analysis of BrdU-tagged cells. BrdU tags PDG cells (arrow) initially, then tagged cells migrate out into duct epithelium, and some of them are shed into the lumen (arrowhead). Double staining of BrdU and Ki67 shows that double-positive cells were found only on day 0. The intestine show the typical dynamics of BrdU positive cells, which migrate from the crypt toward the surface then disappear. (E) Quantitative analysis of BrdU-positive cells shows a dynamic shift of BrdU-positive cells from PDG to the ductal epithelium. (F) A few BrdU-positive cells (arrow) remain in PDG for up to 21 days. Original magnification × 200 (except for intestine; × 100).

Figure 2. TFF are expressed in injured PDG

(A, B) Immunohistochemistry of TFF1 and TFF2 in mouse and human specimens. In normal pancreata, TFF1 is expressed at a low level and identified in only a few cells throughout the ductal epithelium and PDG compartment (arrow). Similarly, TFF2 is expressed at a very low level. In response to inflammatory injury, both TFF1 and TFF2 expression is significantly up-regulated. While TFF1 expression is found in both PDG and duct epithelium (arrowhead), TFF2 expression distribution is restricted to the PDG compartment. Original magnification x400 (mouse), x200 (human) (C, D) Quantitative analysis of TFF-positive PDGs in mouse and human specimen. Both TFF1 and TFF2 are expressed significantly higher in injured pancreas (TFF1: 18.8% and 87.7% (mouse), 19.4% and 84.7% (human); TFF2: 55.0% and 80.9% (mouse), 54.8% and 74.2% (human) in control and pancreatitis, respectively). Data are shown as means +/− SD. (* p<0.01)

Figure 3. Migration of pancreatic ductal epithelial cells is regulated by TFF

(A) Qualitative analysis of in vivo pulse-chase experiments on TFF2-KO mice. BrdU-positive cells remain in the PDG (arrow) compartment over the chase period of 7 days. Ulceration of the pancreatic duct can be seen on day 5 and day 7 (arrowhead). (B) Quantitative analysis shows no dynamic shift of BrdU-tagged cells. (C) HPDE cells were transfected with control and TFF1-expression vector. The transcriptions of TFF1 were confirmed by RT-PCR. (D) TFF1-HPDE cells migrate into artificial wound more rapidly than control cells. (E) Quantitative analysis of the “cell-empty area.” TFF1-transfected cells resulted in a significantly smaller empty area (53.8% vs 29.6%, * p<0.01). (F) Proliferative activity of HPDE cells by resazurin assay. TFF1-transfected cells show no difference in proliferative activity.

Figure 4. PDG give rise to differentiated progeny

Identification of molecular characteristics of BrdU-tagged cells before migration (day 0, arrow) and after migration (day 5, arrowhead) in mouse. During their migration, BrdU-tagged cells lose their initial expression of PDG-specific markers (Alcian blue, TFF2 and Muc6) but acquire the expression of a mature pancreatic duct-specific marker, KRT7. Original magnification ×400.

Figure 5. Micorarray analysis of RNA from PDG and main duct epithelium of normal pancreas

(A) Gene set enrichment analysis (GSEA) using the 424 genes up-regulated in PDG demonstrates significant enrichment of Sox2 and Nanog target genes, as well as genes with altered expression in pancreatic cancer. The gene set name, number of PDG up-regulated genes in the gene set, and p-value are shown. (B) Venn diagram comparing genes enriched in PDG and the crypts of the intestine and colon. (C) Heatmap representation of genes enriched in PDG that are also enriched in intestinal and/or colon crypts. Genes implicated in the biology of stem or progenitor cells are denoted by an asterisk. (D) The expression of SOX2 and nanog in PDG compartment were confirmed within PDG (arrow). Inlet shows higher magnification of positive cells.

Figure 6. PDG and GI stem cell niches share similar organization and function

Comparisons between PDG (right) and the antral glands of the stomach (left).